Environmental Migration

Local-level displacement data are available from humanitarian agencies engaged in providing relief to persons who leave their place of habitual residence due to sudden or progressive changes in the environment. For example, IOM’s DTM is being used to track and monitor disaster displacement (IOM, 2023c). It is designed to regularly and systematically collect information on the movements and evolving needs of displaced persons in both natural disaster and conflict settings. Data is collected at the group/location level through key informant interviews and direct observation or at the household/individual level through surveys and direct interviews with the affected household/individuals. In a recent thematic paper, DTM has provided recommendations for leveraging the use of its data collection for studying the link between human mobility and environmental degradation, climate change and natural disasters, which might be extended to other data collection on the movement and needs of displaced populations (IOM, 2020a).

Because environmental factors are oftentimes linked to other drivers of mobility and not necessarily the initial cause of displacement, it is recommended that questions directed at informants or migrants allow for multiple reasons for migration, for example by including a ranking scale system.
A follow-up question for those who cited natural disasters as one of the reasons for movement would allow for distinguishing between different types of natural disasters such as floods, drought, fires and landslides.
Including questions on the factors preventing certain groups from migrating would address situations where environmental changes act as a barrier to out-migration, return or reintegration, for example in the case of “trapped populations” or those who do not migrate but are at risk of having to stay behind in places where they are more vulnerable to environmental shocks (IOM, 2019).
Questions on income streams and livelihood practices may shed light on how households are affected by environmental changes, especially if incomes and livelihoods are primarily agricultural. Follow-up questions on changes to these livelihood activities would more directly capture the impact of agricultural changes. Similarly, questions related to changes in the availability of resources such as water, timber, farmland and grazing land, might also be included.
Persons affected by environmental changes might also adopt certain coping or adaptation strategies that might be reflected in direct interviews, such as increasing seasonal migration, alternating crops, moving house or relocating to a new area, using different building materials, or changing eating or spending habits.
Linking information on (movement caused by) environmental changes to data on economic migration, particularly seasonal migration, would provide greater insights into environmental migration. Through consistent monitoring of pastoralist movements, DTM’s Transhumance Tracking Tool (TTT), allows for highlighting unusual movements that might be related to environmental changes, potentially serving as an early warning system/alert mechanism (IOM, 2020b).
The relationship between environmental factors and mobility can also be studied by overlapping meteorological data such as average rainfall, average temperatures and fires with Geographic Information System maps of recent displacements.

DTM sometimes uses high-definition images from Geographic Information Systems to identify areas with populations at risk of being displaced by natural disasters, such as houses that are close to watersheds or discharge areas. DTM also uses drone images from before and after natural disasters struck to inform humanitarian relief operations. Drone images can also be used to identify geographic areas for conducting for a household survey on environmental migration (IOM, 2016).

The Internal Displacement Monitoring Centre (IDMC) has compiled data on internal displacement in the context of disasters since 2008 through its online Global Internal Displacement Database (GIDD) (https://www.internal-displacement.org/database/displacement-data/). The estimates are based on information provided by national authorities, UN agencies, non-governmental organizations and the media. Figures are published in the annual Global Report on Displacement (GRID), which also includes figures on internal displacement caused by conflict and generalised violence (IDMC, 2023). IDMC is developing methodologies to map and assess future disaster displacement risks and is starting to gather data on cross-border disaster displacement.

Data on environmental displacement can also be gleaned from humanitarian visas which are issued by countries such as the US, Argentina, Brazil, Ecuador and Mexico.

Big data generated by mobile phone users after disasters such as the 2010 earthquake in Haiti and several typhoons in Philippines and Bangladesh can be used to locate environmentally displaced persons in need of assistance. (IOM, 2016).

Data strengths and limitations

Environmental degradation is particularly challenging to disentangle from other drivers of human mobility and might actually be an underlying cause of conflict-driven displacement (IOM, 2020a). This makes data collection on environmental factors beyond natural disasters leading to evacuations more challenging (IOM, 2023c). Furthermore, most studies are qualitative and focused on specific regions or provinces, which prevents cross-national comparisons of environmental migration (IOM, 2016). There are, however, some secondary sources providing cross-national comparative data, such as the IDMC’s Global Internal Displacement Database. Even when data such as those from the Global Internal Displacement Database are available, these data do not capture the duration of displacement, which is crucial for studying people in long-term displacement, or so-called protracted situations (IOM, 2023c).

Cost-effective ways of generating new data on policy-relevant migrant groups

Figure 1:

Adding a module or relevant questions to existing data sources

This might involve adding questions enabling the identification of different migrant groups to a multi-topic national survey or adding questions on the relevant demographic and socio-economic characteristics to an existing survey which already facilitates the identification of different migrant groups. For example, the range of questions that can be studied using traditional survey data could be extended by taking some simple steps, such as including a basic migration module for all household members as part of the household roster (de Brauw and Carletto, 2012). There are several advantages to adding to an existing survey: they are regularly conducted across many countries and include questions on a range of relevant individual- and contextual-level characteristics; they are generally designed and administered by NSOs which means they are more likely to adhere to statistical standards and to be representative of the general population; the protocols for dissemination of the data have already been established.

Data integration

Combining data is a cost-effective way of arriving at new empirical conclusions. For example, the location or place of residence of respondents in censuses and surveys can linked to map-based information such as the presence of health, education, and employment services, enabling an analysis of the contextual factors linked to migration. Combining and repurposing different types of data can enhance the impacts of data on development. Development problems are complex, spanning economic, cultural, environmental, demographic, and many other factors. Policy design based on data covering only one factor will be incomplete, and sometimes ill-advised. Combining different types of data can fill data gaps and offer new perspectives on development problems (World Bank, 2021).

Data linkages

Linkages can also be made for the same individual across multiple data sources. For example, a census can be used to develop a longitudinal perspective by linking it to other data sources. This would be particularly useful for studying the trajectories of refugee and refugee-like populations (EGRISS, 2018). If a unique PIN is available (and there are protocols and procedures for sharing and disseminating anonymized data) then multiple administrative data sources can be linked, providing more insights not only into the demographic and socio-economic characteristics of migrants but also into any changes they may have experienced over time. There are, however, two ways of linking individual data from multiple sources. Record linking, or linking information on identical units, requires a unique PIN. In the absence of a PIN, statistical matching, or linking based on characteristics that are more likely to identify the same individual, i.e., a combination of employer, birthday, and place of residence, can be used as an alternative (EGRISS, 2018).

Using existing data sources as a sampling frame

The proportions of migrants in population censuses and household surveys tend to be rather small, as these populations are often under-sampled. However, the locations of any migrants identified in these data sources can used for designing sampling frames for specialized migration surveys. As illustrated in the previous chapter, spatial data on new camps and settlements can be used to identify the locations of forcibly displaced populations (EGRISS, 2018).

Chapter overview